Frequently Asked Question
Can you please provide a detailed description of how partial stroke testing is used in the calculation of SFF. In particular, what is the basis for ignoring PVST(manual) in the calculation of SFF but taking credit for PVST(automatic) at the same interval?
Last Updated 5 years ago
When you select PVST, regardless if you set it to manual or automatic, the PFDavg calculation will consider the interval you specified, i.e. there is no difference in PFD algorithm. The only difference is indeed in the SFF calculation. The basis for this is that if the test is executed automatically, we can "guarantee" that it is done, whereas if the test is executed manually, we would need to take the human performance/failure rate into consideration to determine what the level of guarantee is that the test will be done. The latter we do not take into consideration at this point. Therefore the statement, that in order to get credit for an automatic diagnostic, that diagnostic should be automatic.
Why not account for this in the PFD calculation, that deals with probability over a long period of time, whereas the SFF is a single on off number.
It is not assumed that the plant will be taken to the safe state automatically, rather that the failure is detected and that either a repair action is initiated (within the assumed MRT) or the plant is transitioned to the safe state (maybe through use of a manual valve).
Static refers to the frequency of the partial stroke test. If the test is done at least once every 7 days (and is initiated automatically) the test will be classified as a dynamic test. What this means is that certain failure mechanisms like stiction will not occur and therefore the failure behaviour of the solenoid for example will be different. If we have data to support this in the SERH, we will then use the dynamic data rather than static data.
Why not account for this in the PFD calculation, that deals with probability over a long period of time, whereas the SFF is a single on off number.
It is not assumed that the plant will be taken to the safe state automatically, rather that the failure is detected and that either a repair action is initiated (within the assumed MRT) or the plant is transitioned to the safe state (maybe through use of a manual valve).
Static refers to the frequency of the partial stroke test. If the test is done at least once every 7 days (and is initiated automatically) the test will be classified as a dynamic test. What this means is that certain failure mechanisms like stiction will not occur and therefore the failure behaviour of the solenoid for example will be different. If we have data to support this in the SERH, we will then use the dynamic data rather than static data.